While all the determinants of tropism of HIV replication are still being elucidated, the major determinants reside in the envelope of the virus and are present in both the extracellular component (gp120) and the transmembrane component (gp41). However, other genes (gag, nef, vpr, vif) and cis-acting elements (the long terminal repeat, LTR) can also modulate replication in different cell types. Most primary isolates, particularly those of the NSI (non-syncytium inducing) class, fail to replicate in established CD4-positive cell lines, although peripheral blood mononuclear cells (PBMC) are almost always permissive even for NSI viruses. We have shown that the ELI strain of HIV-1 adapts to replicate in H9 and U937 cells, and the changes that confer this enhanced infectivity and expanded host range to this virus are in both gp120 and gp41 - in the CD4-binding region of gp120 and the fusogenic region of gp41. Concomitant with the increased infectivity was an increased ability of the virion to bind to CD4, although CD4 binding to the variant soluble gp120 was not affected. These results revealed that alterations in either gp120 or gp41 in the context of the virion can affect both the capacity of the envelope to interact with CD4 as well as infectivity, thus indicating the importance of these regions to the functional association of the two envelope components. This analysis is being extended using site-directed mutagenesis of these regions in Env. Efficient particle release in HIV-1 is mediated at least in part by the auxiliary gene vpu. Despite not having a corresponding gene, HIV-2 particles are released from the cell as efficiently as those of HIV-1. To map the viral determinants in HIV-2 that facilitate this activity, a hybrid virus was constructed between HIV-1 and HIV-2. The efficiency of particle release was assessed by comparing the proportion of virion-associated and secreted Gag compared with the total Gag produced. Replacing the gag/pol gene of HIV-1 with the HIV-2 gag/pol gene had no effect on particle release. In a Vpu mutant background, particle release was reduced in both viruses and supply of Vpu by co-transfection corrected the defect. When the HIV-1 envelope gene was mutated and the HIV-2 envelope supplied by co-transfection, the amount of particle release was restored and no effect of Vpu could be detected. Therefore, for HIV-2, determinants of efficient particle release map to the HIV-2 env gene and a separate protein, Vpu, is not required as it is for HIV-1.